knitr::opts_chunk$set(warning = FALSE)BMIN503 Final Project Draft: Kole Buckley
Overview:
The overarching goal of this project is to explore potential mechanisms for an increased risk of gastric cancer among BRCA1 and BRCA2 pathogenic germline variant (PV) carriers. Herein, we detail one route of exploration involving the use of BRCA1 and BRCA2 PV patient-derived gastric organoids (PDGOs). Using PDGOs we look at differences in the number of organoids formed, growth, and morphology between PDGOs generated from BRCA1 and BRCA2 PV carriers as well as compared to PDGOs derived from patients without a known gene mutation.
Introduction:
Breast cancer susceptibility gene one (BRCA1) and two (BRCA2), are well established tumor suppressor genes that play a pivotal role in promoting homologous recombination in response to DNA damage. Indeed, pathogenic germline variants (PGVs) in either of these genes leads to an increase in non-homologous end joining, which can in turn promote genomic instability and tumorigenesis. While BRCA1/2 PV carriers are well established to be at risk for breast and ovarian cancer, there is mounting evidence that BRCA1/2 PGV carriers also have an elevated risk of gastric cancer (GC). Indeed, a recent study, among others, showed a cumulative risk as high as 21.3% for BRCA1 and 19.3% for BRCA2 PV carriers by age 85, however this study was performed in a Japanese population where risk of GC is increased at baseline. Importantly, despite this potential increased GC risk, there are currently no recommended GC surveillance guidelines for BRCA1/2 PGV carriers. Furthermore, mechanistic implications of gastric carcinogenesis remain obscure as little to no studies have attempted to explore this in the context of a germline BRCA1/2 PGV.
To begin to elucidate mechanisms associated with GC carcinogenesis in BRCA1/2 PGV carriers we utilized patient-derived gastric organoids (PDGOs) generated from gastric biopsies acquired during an upper endoscopic procedure. Using PDGOs we compare the number of organoids formed, growth, and morphology between BRCA1 and BRCA2 PGV carriers as well as those from patients without a known mutation. This work intersects with several different fields, including gastroenterology, genetics, and microbiology. The goal of this work is to identify a mechanistic avenue to further explore in future studies.
Methods:
Tissue Acquisition:
This research has been approved by the Institutional Review Board (IRB) at the University of Pennsylvania. All human tissue utilized in this protocol was collected from individuals who provided informed consent for tissue collection. Patient gastric biopsies were obtained from the Hospital of the University of Pennsylvania via upper endoscopy. For each patient, gastric biopsies were take from both the body and antral regions of the stomach. In total biospies were collected from both the gastric body and antrum from 10 different BRCA1 PGV carriers, 6 BRCA2 PGV carriers, and 8 control patients. These biopsies were used to generate PDGOs and which were analyzed separately based on the location of the biopsy (body vs. antrum).
Patient-derived Gastric Organoid Generation and Data Collection:
Gatric biopsy tissue from individual patients were enzymatically digested to liberate single cells. Single cells were embedded in a basement membrane extract and seeded in 24-well cell culture plates at a standardized density of 100,000 cells per a well. The entirety of each well was imaged at 10-, 15-, and 20-days post initiation. From these images, organoid number, size, and morphology were measured using an image analysis software (cellSens, Olympus). Data were imported and organized using Excel.
Importing the Data Set from Excel:
Here I import the data set from an Excel file. I do not do any “cleaning” as I had already organized this data in a comprehensive manner as I collected it, prior to this project.
# intall readxl, if needed
#install.packages("readxl")
# load readxl
library(readxl)
# import my exel dataset
excel_data_path <- "/Users/kolebuckley/Desktop/BMIN_503/final_project/Data/final_project_data.xlsx"
excel_data <- read_excel(excel_data_path)
# here i show the first 5 rows as an example of the what the data looks like
head(excel_data, n = 5)# A tibble: 5 × 27
Sample Location Genotype Organoid_num_D10 Area_D10 Perimeter_D10 Radius_D10
<chr> <chr> <chr> <dbl> <dbl> <dbl> <dbl>
1 GC066D Antral BRCA1 70 9345. 399. 42.1
2 GC067P Body BRCA2 30 2833. 239. 28.6
3 GC067D Antral BRCA2 86 3995. 308. 32.3
4 GC070P Body CTL 22 5420. 278. 37.4
5 GC070D Antral CTL 22 8531. 346. 44.1
# ℹ 20 more variables: ShapeFactor_D10 <dbl>, Sphericity_D10 <dbl>,
# Elongation_D10 <dbl>, Diameter_D10 <dbl>, Organoid_num_D15 <dbl>,
# Area_D15 <dbl>, Perimeter_D15 <dbl>, Radius_D15 <dbl>,
# ShapeFactor_D15 <dbl>, Sphericity_D15 <dbl>, Elongation_D15 <dbl>,
# Diameter_D15 <dbl>, Organoid_num_D20 <dbl>, Area_D20 <dbl>,
# Perimeter_D20 <dbl>, Radius_D20 <dbl>, ShapeFactor_D20 <dbl>,
# Sphericity_D20 <dbl>, Elongation_D20 <dbl>, Diameter_D20 <dbl>Results:
Representative Images:
Here we show representative images of organoids at 10-, 15-, and 20-days post initiation for BRCA1 and BRCA2 PGV carriers as well as controls (Figure 1).
Figure 1: Representative images of organoid growth from day 10 to 20 post initiation. Organoids shown were derived from the gastric body. Scale bars = 500 um.
Gastric Organoid Number:
We first looked to see if there are differences in the number of organoids generated from gastric biopsies of the gastric body and antrum among BRCA1 and BRCA2 PGV carriers as well as controls (no PGVs). We measure organoid number at 10-, 15-, and 20-days post initiation.
# here i create bar graphs to show differences betwen organoid number at 10, 15, and 20 days post organoid initiation between organoids from BRCA1 carriers, BRCA2 carriers, and control
# load necessary packages
library(ggplot2)
library(tidyverse)
# Filter the data so as to only examine the organoids dervied from biopsies of the gastric body first
filtered_data_body <- excel_data %>%
filter(Location == "Body")
# Reshape the data to long format for easier usage
tidy_data_body <- filtered_data_body %>%
gather(key = "Timepoint", value = "OrganoidNumber", Organoid_num_D10, Organoid_num_D15, Organoid_num_D20)
# Create a bar plot
ggplot(tidy_data_body, aes(x = Timepoint, y = OrganoidNumber, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Number of Organoids Derived From Biopsies of the Gastric Body",
y = "Number of Organoids",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
# for organoids dervied from biopsies of the antral region of the stomach
filtered_data_antral <- excel_data %>%
filter(Location == "Antral")
tidy_data_antral <- filtered_data_antral %>%
gather(key = "Timepoint", value = "OrganoidNumber", Organoid_num_D10, Organoid_num_D15, Organoid_num_D20)
ggplot(tidy_data_antral, aes(x = Timepoint, y = OrganoidNumber, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Number of Organoids Derived From Biopsies of the Gastric Antrum",
y = "Number of Organoids",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
From this data we see that organoids derived from both BRCA1 and BRCA2 PGV carriers grow more numerous than those derived from controls, at all timepoints. While there does not appear to be any differences in the number of organoids between BRCA1 and BRCA2 PGV carriers when organoids are generated from gastric body, it does appear that BRCA2 organoids generated from the gastric antrum grow more numerous than BRCA1.
Gastric Organoid Growth:
We next looked at the size of organoids generated from gastric biopsies of the gastric body and antrum among BRCA1 and BRCA2 PGV carriers as well as controls. This was done by aquiring images of the entire well in which the organoids grew in and analyzed via an image analysis software. Organoid size is depicted as area (um^2) *not sure how add superscript in R!
# here i create a bar graphs to show differences betwen organoid size as measured by area at 10, 15, and 20 days post organoid initiation between organoids from BRCA1 carriers, BRCA2 carriers, and control
# Reshape the data to long format for easier usage, we will look at organoid area derived from biopsies of the gastric body first
tidy_data_body_size <- filtered_data_body %>%
gather(key = "Timepoint", value = "OrganoidSize", Area_D10, Area_D15, Area_D20)
# Create a bar plot
ggplot(tidy_data_body_size, aes(x = Timepoint, y = OrganoidSize, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Area of Organoids Derived From Biopsies of the Gastric Body",
y = "Organoid Area (um^2)",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
# for organoids derived from biopsies of the gastric antrum
tidy_data_antral_size <- filtered_data_antral %>%
gather(key = "Timepoint", value = "OrganoidSize", Area_D10, Area_D15, Area_D20)
ggplot(tidy_data_antral_size, aes(x = Timepoint, y = OrganoidSize, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Area of Organoids Derived From Biopsies of the Gastric Antrum",
y = "Organoid Area (um^2)",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
As it relates to organoids derived from the gastric body, it appears that organoids from BRCA1 PGV carriers and controls grow at a similar pace, as measured by organoid area. Although, by day 20 controls are the largest. Meanwhile, body organoids from BRCA2 PGV carriers show little growth. For organoids derived from the gastric antrum, we see that those from BRCA1 PGV carriers exhibit a slightly greater growth compared to BRCA2 PGV carriers and controls.
Next we will look at organoid diameter.
#here i create a bar graphs to show differences betwen organoid size as measured by diameter at 10, 15, and 20 days post organoid initiation between organoids from BRCA1 carriers, BRCA2 carriers, and control
# Reshape the data to long format for easier usage, we will look at organoid diameter derived from biopsies of the gastric body first
tidy_data_body_diam <- filtered_data_body %>%
gather(key = "Timepoint", value = "OrganoidDiameter", Diameter_D10, Diameter_D15, Diameter_D20)
# Create a bar plot
ggplot(tidy_data_body_diam, aes(x = Timepoint, y = OrganoidDiameter, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Diameter of Organoids Derived From Biopsies of the Gastric Body",
y = "Organoid Diameter (um)",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
#for organoids derived from biopsies of the gastric antrum
tidy_data_antral_diam <- filtered_data_antral %>%
gather(key = "Timepoint", value = "OrganoidDiameter", Diameter_D10, Diameter_D15, Diameter_D20)
ggplot(tidy_data_antral_diam, aes(x = Timepoint, y = OrganoidDiameter, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Diameter of Organoids Derived From Biopsies of the Gastric Antrum",
y = "Organoid Diameter (um)",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
As can seen been here, organoids diameter largely follows the same trends seen when measuring organoid area. However for organoids derived from the gastric antrum, there is a bit more separation for organoids for BRCA1 PGV carriers and controls compared to BRCA2 PGV carriers.
Lastly we look a organoid perimeter.
#here i create a bar graphs to show differences betwen organoid size as measured by perimeter at 10, 15, and 20 days post organoid initiation between organoids from BRCA1 carriers, BRCA2 carriers, and control
# Reshape the data to long format for easier usage, we will look at organoid perimeter derived from biopsies of the gastric body first
tidy_data_body_perim <- filtered_data_body %>%
gather(key = "Timepoint", value = "OrganoidPerimeter", Perimeter_D10, Perimeter_D15, Perimeter_D20)
# Create a bar plot
ggplot(tidy_data_body_perim, aes(x = Timepoint, y = OrganoidPerimeter, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Perimeter of Organoids Derived From Biopsies of the Gastric Body",
y = "Organoid Perimeter (um)",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
#for organoids derived from biopsies of the gastric antrum
tidy_data_antral_perim <- filtered_data_antral %>%
gather(key = "Timepoint", value = "OrganoidPerimeter", Perimeter_D10, Perimeter_D15, Perimeter_D20)
ggplot(tidy_data_antral_perim, aes(x = Timepoint, y = OrganoidPerimeter, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Perimeter of Organoids Derived From Biopsies of the Gastric Antrum",
y = "Organoid Perimeter (um)",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
As can seen been here, organoids perimeter largely follows the same trends seen when measuring organoid area and diameter. However, like the unlike the area findings, organoids derived from the gastric antrum show a bit more separation for organoids from BRCA1 PGV carriers and controls compared to BRCA2 PGV carriers, when measuring perimeter.
Gastric Organoid Morphology:
Organoid morphology as measured by sphericity (where a value of 1 equals a perfect sphere) was accessed for BRCA1 and BRCA2 PGV carriers as well as controls.
For any given well of organoids, we saw a diversity of morphologies. An example of different organoid morphologies are shown in Figure 2.
Figure 2: Organoid morphology. Scale bars = 200 um.
# here i create a bar graph to show differences betwen organoid morphology as measured via Sphericity at 10, 15, and 20 days post organoid initiation between organoids from BRCA1 carriers, BRCA2 carriers, and control
# Reshape the data to long format for easier usage, we will look at organoid sphericity derived from biopsies of the gastric body first
tidy_data_body_morph <- filtered_data_body %>%
gather(key = "Timepoint", value = "OrganoidMorphology", Sphericity_D10, Sphericity_D15, Sphericity_D20)
# Create a bar plot
ggplot(tidy_data_body_morph, aes(x = Timepoint, y = OrganoidMorphology, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Spheriicity of Organoids Derived From Biopsies of the Gastric Body",
y = "Organoid Sphericity",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
# for organoids derived from the gastric antrum
tidy_data_antral_morph <- filtered_data_antral %>%
gather(key = "Timepoint", value = "OrganoidMorphology", Sphericity_D10, Sphericity_D15, Sphericity_D20)
ggplot(tidy_data_antral_morph, aes(x = Timepoint, y = OrganoidMorphology, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Sphericity of Organoids Derived From Biopsies of the Gastric Antrum",
y = "Organoid Sphericity",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
Despite observing a variety of morphologies in a given well, on average, the results here show little to no differences in sphericity amongst BRCA1 and BRCA2 PGV carriers as well as controls.
Next we will look at organoid shape factor (where a value of 1 is equal to a circle with an equal perimeter). ***May not include this data in the final draft as it is a somewhat similar measurement to sphericity.
# here i create a bar graph to show differences betwen organoid morphology as measured via Shape Factor at 10, 15, and 20 days post organoid initiation between organoids from BRCA1 carriers, BRCA2 carriers, and control
# Reshape the data to long format for easier usage, we will look at organoid shape factor derived from biopsies of the gastric body first
tidy_data_body_sf <- filtered_data_body %>%
gather(key = "Timepoint", value = "OrganoidShapeFactor", ShapeFactor_D10, ShapeFactor_D15, ShapeFactor_D20)
# Create a bar plot
ggplot(tidy_data_body_sf, aes(x = Timepoint, y = OrganoidShapeFactor, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Shape Factor of Organoids Derived From Biopsies of the Gastric Body",
y = "Organoid Shape Factor",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
# for organoids derived from the gastric antrum
tidy_data_antral_sf <- filtered_data_antral %>%
gather(key = "Timepoint", value = "OrganoidShapeFactor", ShapeFactor_D10, ShapeFactor_D15, ShapeFactor_D20)
ggplot(tidy_data_antral_sf, aes(x = Timepoint, y = OrganoidShapeFactor, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Shape Factor of Organoids Derived From Biopsies of the Gastric Antrum",
y = "Organoid Shape Factor",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
While there are some slight changes to this data compared to organoid sphericity, the overall trends are similar. That is, there does not appear to be any differences among organoids derived from BRCA1 and BRCA2 PGV carriers and controls when evaluated via shape factor.
Lastly, we look at organoid elongation (the higher the value, the more elongated the organoid is shaped).
# here i create a bar graph to show differences betwen organoid morphology as measured via elongation at 10, 15, and 20 days post organoid initiation between organoids from BRCA1 carriers, BRCA2 carriers, and control
# Reshape the data to long format for easier usage, we will look at organoid elongation derived from biopsies of the gastric body first
tidy_data_body_elong <- filtered_data_body %>%
gather(key = "Timepoint", value = "OrganoidElongation", Elongation_D10, Elongation_D15, Elongation_D20)
# Create a bar plot
ggplot(tidy_data_body_elong, aes(x = Timepoint, y = OrganoidElongation, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Elongation of Organoids Derived From Biopsies of the Gastric Body",
y = "Organoid Elongation",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
# for organoids derived from the gastric antrum
tidy_data_antral_elong <- filtered_data_antral %>%
gather(key = "Timepoint", value = "OrganoidElongation", Elongation_D10, Elongation_D15, Elongation_D20)
ggplot(tidy_data_antral_elong, aes(x = Timepoint, y = OrganoidElongation, fill = Genotype)) +
geom_bar(stat = "summary", fun.y = "mean", position = "dodge") +
labs(title = "Average Elongation of Organoids Derived From Biopsies of the Gastric Antrum",
y = "Organoid Elongation",
x = "Timepoint") +
theme_minimal() +
scale_fill_manual(values = c("blue", "black", "gray"))
Again, there are minute differences between this data and that for sphericity and shape factor, however, the overall trends look similar. That is, there does not appear to be any differences among organoids derived from BRCA1 and BRCA2 PGV carriers and controls when evaluated via elongation.
Discussion:
Currently working on this..